64

International Journal of Plasma Environmental Science & Technology, Vol.7, No.1, MARCH 2013

The Influence of the Electrophoretic and Polarization Forces on Two Phase

Flow Redistribution in a Horizontal Annular Tube
S. Nangle-Smith, H. Sadek, and J. S. Cotton
Department of Mechanical Engineering, McMaster University, Canada
AbstractThe objective of this study is to investigate the effect of high voltage DC waveforms on two phase flow
redistribution. The flow patterns in a two phase flow system directly influences the system heat transfer and pressure
drop, and by extension, the system performance. Electrohydrodynamics (EHD) induces flow pattern redistribution and
therefore can provide a low power, rapid method of enhancement of two phase flow systems such as heat exchangers. The
heat transfer characteristics can be varied by modifying the proportion of liquid or vapor in contact with the heat transfer
surface. This study investigates the effect of the electrophoretic and polarization components of the EHD force on two
phase flow redistribution in an annular channel for comparable cases of condensation and evaporation.
KeywordsElectrohydrodynamics, two-phase flow, polarity, electrophoretic, polarization

I. INTRODUCTION
The EHD body force term in the fluid momentum
equations causes flow redistribution. For dielectric two
phase flows, the force is composed of three terms as
shown in equation (1).
1
1
f eB ei E E 2 E
2
2

(1)

The first term is known as the electrophoretic or

Coulomb force. This force acts on free ions or charges
within the fluid and causes them to migrate along electric
field lines, a phenomenon known as electroconvection
[1]. The electrophoretic force is dominant in single phase
applications, where there is little change in the
permittivity, , of the fluid. In two phase applications
there will be electroconvection currents in both the liquid
and vapor phases due to this force. This force component
is proportional to the electric field, E, unlike the other
two forces which are proportional to E2. Therefore the
effect of polarity manifests in this force. The
electrophoretic force is proportional to the volume charge
density, ei, in the fluid and so charge injection plays a
role. For refrigerant systems similar to our design, Ng [2]
suggested that the mechanism of charge injection is
always negative charges from the negative electrode
based on the mobility model proposed by Fujino et al. [3].
The second and third terms of equation (1) are known
collectively as the polarization forces. They polarize or
stretch the molecules in the dielectric. The
dielectrophoretic force arises due to the spatial change in
permittivity. The permittivity can be spatially
inhomogeneous due to non-uniform electric fields,
changes in phase, temperature, T, and density, ,
gradients in the flow. The electrostrictive term arises due
Corresponding author: James S. Cotton
e-mail address: cottonjs@mcmaster.ca
Presented at the 2012 International Symposium
Electrohydrodynamics (ISEHD 2012), in September 2012

to the change in permittivity with density which can be

due to local concentrations (i.e. fluid additives,
suspended droplets, boundary layers etc.) or temperature
gradients in the flow. Large gradients in permittivity
exist at the liquid-vapor interface for two phase systems.
The polarization forces as a result of this interfacial stress
are larger than the electrophoretic force by an order of
magnitude [4]. Phase redistribution occurs due to the
difference in relative permittivity between the phases; 9.5
for the R134a liquid and approximately 1 for the R134a
vapor. The liquid phase is attracted to regions of high
electric field due to its higher permittivity, a phenomenon
known as liquid extraction [1]. Fig. 1 summarizes the
effect of EHD forces on various components in a flow
system.
The main influences on flow pattern in a system are
the system geometry (including objects that disturb the
flow), orientation (gravitational effects), temperature
dependent fluid properties, surface tension in particular,
the flow rates and the proportion of liquid and vapor in

Fig. 1. Effect of EHD on flow redistribution.

on

Nangle-Smith et al.

65

Fig. 2. Convective boiling (top) & convective condensation (bottom) horizontal flow patterns (adapted from Collier [15])

Fig. 3. EHD flow patterns

the system, i.e. the flow quality, x. The flow patterns for
convective boiling and convective condensation in a
horizontal tube are shown schematically in Fig. 2. The
main difference between the two is that an annular flow
regime exists at higher vapor qualities in condensation
than for evaporation. EHD can induce new flow patterns,
e.g. twisted liquid cones and columns, as shown in Fig. 3
[2, 5].
Flow pattern maps have been developed as a
predictive design tool. Theoretically based maps are
recommended as they are applicable to all types of flow
systems, e.g. the Taitel and Dukler general flow pattern
map [6], the El Hajal et al. map for convective
condensation [7] and Kattans map [8] for convective
boiling. The Steiner map [9] incorporates the effect of an
annular geometry on flow patterns. Cotton [4] developed
an EHD flow pattern map which incorporates the EHD
induced interfacial force into the Steiner analytical model
to predict the transition from stratified to annular flow. It
was found that when using EHD this transition can occur
at lower flow rates.

Most convective boiling studies use long test

sections, 1-3 m in length [1, 4, 10]. Although these
lengths are typically those found for evaporators in
industry, it is impossible to maintain a constant flow
pattern when the quality varies so greatly across the
length. Since EHD is known to change the flow pattern
distribution and experimental parameters which quantify
the effects of EHD, e.g. heat transfer coefficient, are
highly dependent on specific flow patterns, it is
necessary to maintain a consistent flow pattern to gain
insight into the true effects of EHD. Some researchers
that have used short, 300mm, test sections for EHD
convective boiling [12-14]. In this study, a short test
section is used to maintain a consistent flow pattern
along the tube and to determine the effect of DC high
voltage on convective boiling and condensation flow
redistribution.
II. METHODOLOGY
Convective boiling and condensation tests subject to
electrohydrodynamics (EHD) were conducted in a
horizontal, counter-current, shell and tube heat exchanger.
The test rig used in this study is the same as that used in
studies by Sadek and Ng [2, 16]. The test section is
shown Fig. 4. The outer tube, inner diameter 10.2 mm, is
grounded and voltage is applied via a concentric stainless
steel electrode, outer diameter 3.2 mm. The test section is
300 mm long allowing for a consistent flow pattern along
the tube length. DC voltages of varying polarity and
amplitude up to 8 kV were tested for the following flow
conditions: Mass flux, G: 60 kg/m2s, Average Quality,
Xav: 40%, inlet and outlet quality difference, X: 10%.
The applied water side heat flux, q, was positive or

66

International Journal of Plasma Environmental Science & Technology, Vol.7, No.1, MARCH 2013

Fig. 4. Test section schematic.

TABLE I
EXPERIMENTAL UNCERTAINTIES
Measurement, x
RTD Twater
Thermocouple temp.
Refrigerant flow rate
Water flow rate
Pressure drop (Low)
Pressure drop (High)

C
C
kg/s
kg/s
Pa
Pa
o

Parameter, f
Heat flux, q
Heat transfer coefficient, h
Inlet quality, Xin

Precision
error
0.08
0.24
1.210-3
210-4
10
18

Bias
error
0.04
0.22
1.310-5
310-4
11
13

Total
0.09
0.33
1.210-3
3.610-4
15
23

Relative uncertainty
kW/m2
W/m2K
dim

12%
13%
23%
within 5%

System energy balance

negative 8-9 kW/m2 for condensation or evaporation

respectively. Table I shows the experimental
uncertainties and the overall system energy balance. The
flow parameters were chosen to maintain a wavy liquid
stratum below the electrode along the test section before
the application of EHD to assess the mechanism of
transition to other flow distributions, similar to the
analysis of Taitel and Dukler [6]. It has been shown that
this flow distribution allows for high interfacial EHD
forces induced by the proximity of the two phase
interface to the electrode for a shell and tube heat
exchanger system [17].
The water side heat flux is calculated as:
"
qwater
m waterCP, waterTwater

(2)

Also, the average heat transfer coefficient is calculated as

hav

"
qwater
As Tref , sat Ts , av

(3)

where is the flow rate, Cp is the specific heat capacity

of the water, Twater is the water temperature difference
measured using a RTD, As is the heat exchanger surface
area, Tref,sat is the saturated refrigerant temperature and
Ts,avg is the average surface temperature.
III. RESULTS AND DISCUSSION
Fig. 5 shows the flow visualization using a Photron
high speed camera and the associated estimated static
electric field distribution for the onset of high voltage
application in a stratified wavy flow system. The electric
field distribution was calculated using Comsol
maintaining a constant void fraction, = 0.164, between
each case, i.e. the same cross sectional area of liquid.
The electric field ranges from 8104 to 3.3106 V/m,
with the darker areas indicating areas with higher electric
field. Fig. 5a suggests that the electric field is strongest at
the bottom of the electrode and thus, the interfacial force
is highest at the midpoint of the interface. This causes the
liquid to be extracted toward the electrode. The
interfacial force increases with decreasing electrodeliquid gap distance and with increasing electric field
strength. If this force is high enough the liquid level will
reach the electrode. From the flow visualization we can
see that liquid extraction is occurring due to this electric
field distribution.
The electric field distribution changes when the
liquid level reaches the electrode. The region of highest
electric field is now at the electrode, either side of the
liquid region. Therefore the liquid tends to be pulled
around the electrode surface due to the polarization
forces at the interface and is possibly partially aided by
surface tension effects. When the liquid has fully
surrounded the electrode, as seen in Fig. 5d, the electric
field is inverted, with the strongest region being at the
liquid-vapor interface offset from the electrode. This will

Nangle-Smith et al.

67

Fig. 6. DC voltage flow visualization - liquid repulsion.

Fig. 7. Transient and steady state EHD patterns.

Fig. 5. DC voltage flow visualization and electric field analysis.

induce a polarization force that repels the liquid away

from the electrode.
The mechanism of liquid repulsion is different for
positive and negative applied voltages and results in
different flow patterns as can be seen in the flow
visualization in Fig. 6. Positive voltages give rise to
twisted liquid conical structures and Negative voltages
induce twisted liquid column-like structures below the
electrode. These differing flow patterns based on polarity
were also seen by Ng [16].
It is suggested here that the electrophoretic force
contributes the most to the difference in repulsion
between positive and negative applied voltage since it is
the component where polarity has a dominant effect. In
the case of negative applied voltage, the electrophoretic
force aids the repulsion force whereas in the case of
positive applied voltage, the electrophoretic force acts
against the repulsion force. For the case of a negative

applied voltage, charge is injected from the negative

electrode liquid annulus surrounding the electrode. The
liquid annulus becomes negatively charged and this
causes it to be repelled from the negative electrode. The
wavy flow on the bottom electrode is not charged unless
there is a gap bridged between this and the liquid
surrounding the electrode. This liquid stratum is
extracted toward the electrode and repelled upon
reaching the electrode. For the case of a positive polarity,
the liquid surrounding the electrode is positively charged
and will be repelled from the positive electrode. However,
charge injection of negative ions occurs from the outer
tube into the wavy flow in contact with this surface
underneath the electrode. This negatively charged liquid
will be highly attracted toward the positive electrode both
due to polarization and electrophoretic forces. When this
liquid is extracted and surrounds the electrode it acts
against the liquid being repelled from the positively
charged electrode. This extraction and surrounding
occurs continually and the proposed result is an inverted
annular flow with some conical regions of liquid
repulsion.
The transient flow patterns occurring milliseconds
after the application of high voltage are shown in Figs. 5
and 6. After approximately 520 ms the electrophoretic
and polarization forces balance out and a steady state will
exist. [16]. Steady state flow patterns are less turbulent
than the transient flow patterns but the differences in
liquid repulsion mechanisms for opposing polarities still
exist as seen in Fig. 7. The transient flow patterns
provide insight into the phenomenon of EHD but the
steady state data, taken after 10 minutes, is used for all
heat transfer and pressure drop readings.

68

International Journal of Plasma Environmental Science & Technology, Vol.7, No.1, MARCH 2013

Fig. 8. Surface temperature profiles at axial location 2. evaporation (top), condensation (bottom).

Fig. 9. Surface temperature associated standard deviations. evaporation (top), condensation (bottom).

Fig. 10. Flow visualization. evaporation (top), condensation (bottom).

The flow redistribution due to EHD can be studied

via flow visualization or by analysis of the surface
temperature profiles and their associated standard
deviations along the test section. The surface
temperatures show are all for location 2 i.e. the middle of
the test section. (See Fig. 4)
For the no EHD evaporation case, the bottom
temperature is approximately 28oC, the top is

approximately 33oC (Fig. 8) and the associated standard

deviation (Fig. 9) is low, 0.06oC approximately. This
suggests a stratified flow pattern with liquid at the
bottom and vapor at the top. The side temperatures vary
between 28oC and 33oC with high variance, 0.2oC
approximately, suggesting a wavy interface that wets the
side. The corresponding flow images in Fig. 10 confirm
this flow pattern. The no EHD condensation case (Fig. 8)

Nangle-Smith et al.

Fig. 11. Heat transfer enhancement comparison: The effect of DC

applied voltage, G = 60 kg/m2s, q = 8.5 kW/m2, Xav = 40%.
Evaporation, Condensation

Fig. 12. Pressure drop comparison: The effect of DC applied voltage,

G = 60 kg/m2s, q = 8.5 kW/m2, Xav = 40%.
Evaporation, Condensation

is quite similar to its evaporation counterpart except the

temperatures are lower, closer in value and all have a low
standard deviation, 0.06oC approximately. Thinner liquid
films in condensation coincide with higher temperatures.
This suggests that liquid film is present at all of the
measurement locations in addition to the stratified flow
pattern and this is confirmed in Fig. 10.
For the case of positive 8 kVDC, Figs. 8 and 9 show
intermittent top and side temperatures in evaporation.
This suggests an intermittent flow with varying liquid
thicknesses. The flow visualization in Fig. 10 shows
conical structures below the electrode, high turbulence,
liquid in core region and some liquid droplets repelled
from the electrode impinging on the heat transfer surface.
A similar flow pattern is seen in Fig. 10 for the
condensation case. The variance is low which may be
due to the condensing liquid film present on the top and
sides surfaces.
Finally for the case of negative 8 kVDC, Figs. 8 and
9 show similar top and side temperatures with low
variance suggesting that there is liquid at these locations.
Fig. 10 confirms an annular flow pattern has been
established due to the liquid repulsion. The same flow
pattern is suggested from Figs. 8 and 10 for the case of
condensation. It seems that the stronger the electric field,

69

the more similar the flow patterns become for

condensation and evaporation. This makes sense as the
effect of EHD becomes more dominant.
The heat transfer coefficient and the test section
pressure drop are used to quantify the enhancement
achieved with EHD. (Figs. 11 and 12 respectively)
Higher voltages result in more flow redistribution and
therefore higher enhancement and a higher pressure drop.
For condensation there is approximately a 2.5 fold
increase in heat transfer for positive applied voltage and
approximately 2.4 fold increase for the negative applied
voltage. The enhancement is approximately 1.5 fold for
positive voltage and 4 fold for negative voltage for
evaporation. See Fig. 11. The pressure drop penalty
increases by approximately 4fold in both cases for the
highest tested applied voltage. See Fig. 12.
The effect of polarity is apparent in this graph
indicating that negative voltages are more suitable for
convective boiling heat transfer enhancement and
positive voltages more suitable for convective
condensation heat transfer. It is hypothesized that heat
transfer and pressure drop are primarily a function of
flow pattern and that some flow patterns may be more
favorable for different heat transfer mechanisms. Here,
the positive applied voltage attracts more liquid into the
core region drying the heat transfer surface. This is
advantageous in condensation, where the vapor needs to
be in contact with heat transfer surface to condense, but
less efficient in boiling as portions of the heat transfer
surface may be dry. The opposite is true for the case of
negative applied voltage. Here more liquid is repelled
toward the heat transfer surface. This is suitable for
boiling as it rewets the heat transfer surface, but not as
useful in condensation as it interferes with the ease of
formation of condensing droplets.
IV. CONCLUSION
A short, 300 mm, test section was used to study the
effect of EHD forces on flow redistribution in a
horizontal, shell and tube heat exchanger subject to both
boiling and condensation. The use of the short test
section allows for a consistent flow pattern across the test
section length which provides further insight into the true
effect of EHD. Flow visualization and surface
temperature measurements were used to study the two
phase flow pattern redistribution under applied DC
voltage.
It was found that polarity influences the flow
distribution. For both polarities, the liquid is extracted
toward the electrode due to the polarization forces acting
at the liquid vapor interface. The flow then surrounds the
electrode via liquid extraction as the liquid, which has a
high dielectric constant, is attracted to the region of high
electric field. Once the fluid surrounds the electrode, the
liquid will be repelled due to the polarization forces
arising from the now inverted electric field distribution.
The electrophoretic force aids this repulsion for negative

70

International Journal of Plasma Environmental Science & Technology, Vol.7, No.1, MARCH 2013

voltages and opposes this force for positive voltages,

generating new flow distributions which were seen via
flow visualization and surface temperature profiles.
A liquid film on the heat transfer surface improves
boiling heat transfer and removal of this film improves
condensation heat transfer. Using this knowledge we
were able to show that these flow distributions could be
used to enhance different modes of heat transfer.

[13]

[14]

[15]
[16]

ACKNOWLEDGMENT
This work is partially sponsored by NSERC. The
author would also like to thank Dr. Cotton, Dr. Ching, Dr.
Sadek and Mr. Ng for their invaluable help throughout
the project.
REFERENCES
A. Yabe, T. Taketani, H. Maki, K. Takahashi, and Y. Nakadai,
"Experimental study of electrohydrodynamically (EHD)
enhanced evaporator for nonazeotropic mixtures," ASHRAE
Transactions, vol. 98, pp. 455-460, 1992.
[2] K. Ng, "Mechanisms of electrohydrodynamic two-phase flow
structures and the influence on heat transfer and pressure drop,"
M. Sc. Thesis, McMaster University, Hamilton, Canada, 2010.
[3] T. Fujino, Y. Yokoyama, and Y. H. Mori, "Augmentation of
laminar forced-convective heat transfer by the application of a
transverse electric field," Journal of Heat Transfer, vol. 111, pp.
345-351, 1989.
[4] J. Cotton, A. J. Robinson, M. Shoukri, and J. S. Chang, "A twophase flow pattern map for annular channels under a DC applied
voltage and the application to electrohydrodynamic convective
boiling analysis," International Journal of Heat and Mass
Transfer, vol. 48, pp. 5563-5579, 2005.
[5] H. Sadek, J. S. Cotton, C. Y. Ching, and M. Shoukri, "Effect of
frequency on two-phase flow regimes under high-voltage AC
electric fields," Journal of Electrostatics, vol. 66, pp. 25-31,
2008.
[6] Y. Taitel and A. E. Dukler, "A model for predicting flow regime
transitions in horizontal and near horizontal gas-liquid flow,"
AIChE Journal, vol. 22, pp. 47-55, 1976.
[7] (a) J. El Hajal, J. R. Thome, and A. Cavallini, "Condensation in
horizontal tubes, part 1: two-phase flow pattern map,"
International Journal of Heat and Mass Transfer, vol. 46, pp.
3349-3363, 2003. (b) J. R. Thome, J. El Hajal, and A. Cavallini,
"Condensation in horizontal tubes, part 2: new heat transfer
model based on flow regimes," International Journal of Heat
and Mass Transfer, vol. 46, pp. 3365-3387, 2003.
[8] (a) N. Kattan, J. R. Thome, and D. Favrat, "Flow boiling in
horizontal tubes: Part 1 - Development of a diabatic two-phase
flow pattern map," Journal of Heat Transfer-Transactions of the
Asme, vol. 120, pp. 140-147, 1998. (b) N. Kattan, J. R. Thome,
and D. Favrat, "Flow boiling in horizontal tubes: Part 3 Development of a new heat transfer model based on flow
pattern," Journal of Heat Transfer-Transactions of the Asme, vol.
120, pp. 156-165, 1998.
[9] D. Steiner, Heat Transfer to Boiling Saturated Liquids, VDIWarmeatlas,Verein Deutscher Ingenieure, ed. VDI-Gessellschaft
Verfahrenstechnik
und
Chemieingenieurwesen
(GCV),
Dusseldorf, Germany, (Translator: J.W. Fullarton), 1993.
[10] A. Singh, M. M. Ohadi, S. Dessiatoun, and M. Salehi, "In-tube
boiling enhancement of R134a utilizing the electric field effect,"
ASME-JSME Papers, vol. 2, pp. 215-223. 1995.
[11] J. S. Cotton, "Mechanisms of Electrohydrodynamic flow and heat
transfer in horizontal convective boiling channels," Ph.D. Thesis,
McMaster University, Hamilton, Canada, 2000.
[12] M. Salehi, M. M. Ohadi, and S. Dessiatoun, "Applicability of the
EHD technique for convective boiling of refrigerant blends [1]

[17]

experiments with R-404A," ASHRAE Transactions, vol. 102, pp.

839-844, 1996.
M. Salehi, M. M. Ohadi, and S. Dessiatoun, "EHD enhanced
convective boiling of R-134a in grooved channels - Application
to subcompact heat exchangers," Journal of Heat TransferTransactions of the Asme, vol. 119, pp. 805-809, 1997.
J. E. Bryan and J. Seyed-Yagoobi, "Influence of flow regime,
heat flux, and mass flux on electrohydrodynamically enhanced
convective boiling," Journal of Heat Transfer-Transactions of
the Asme, vol. 123, pp. 355-367, 2001.
J. G. Collier, Convective Boiling and Condensation, 2nd. s.l. :
McGraw-Hill, New York, 1973.
K. Ng, C. Y. Ching, and J. S. Cotton, "Transient Two-Phase Flow
Patterns by Application of a High Voltage Pulse Width
Modulated Signal and the Effect on Condensation Heat
Transfer," Journal of Heat Transfer-Transactions of the Asme,
vol. 133, 091501, 2011.
H. Sadek, "Electrohydrodynamic control of convective
condensation heat transfer and pressure drop in a horizontal
annular channel," Ph.D. Thesis, McMaster University, Hamilton,
Canada, 2009.